On Wed, 19 Jul 2006, Jes Kasper Klittum wrote:

> 
> > 
> > none of that seems very likely...
> > 
> > can you attach strace to the rdiff-backup process on the remote host?
> > "strace -o tracefile -p $pid"  ... hopefully there'll be some hint at the
> > very end.  if that doesn't help then also try figuring out the ssh
> > processes and shell and tracing those too...
> 
> Attached is the last lines before it died.

assuming that you were tracing the remote rdiff-backup then it appears to 
be getting an EOF on stdin... which makes me think there's something going 
on in the transport layer... not sure what though -- but you could do more 
stracing.

you could also try to figure out the tcp session port and then tcpdump 
that session and look at what happens near the end -- see which side sends 
the FIN first... then go strace the sshd on that side.

this might help <http://arctic.org/~dean/circular.tar.gz> ... it's just a 
buffer i tack on the end of strace outputs when i know there'll be a lot 
of output and i only really need a few MiB of it at the very end to 
diagnose the problem.

-dean




_______________________________________________
rdiff-backup-users mailing list at [email protected]
http://lists.nongnu.org/mailman/listinfo/rdiff-backup-users
Wiki URL: http://rdiff-backup.solutionsfirst.com.au/index.php/RdiffBackupWiki

Reply via email to